Meta Pauses AI Scraping Terms After Publisher Backlash

Meta is delaying the enforcement of its new business terms until November 20, according to The Verge. This comes after publishers raised concerns that the terms might allow Meta to scrape news sites for AI training. The original effective date was October 30.

Meta spokesperson Daniel Roberts stated the delay provides more time to explain the changes to partners and stakeholders. He also noted that content owners can still use robots.txt files to signal their preferences for web crawlers. The terms apply to Meta's business-related products and accounts, which almost every newsroom uses, and there is no opt-out other than ceasing to use Meta's platforms for business.

The original terms, announced in October, allowed Meta to "access, index, cache, analyse, collect, retrieve or crawl" associated domains and URLs. An earlier version explicitly mentioned using collected content to "train" its products. While Meta removed the word "train," legal experts, including media attorney Matt Topic, believe the language remains broad enough to cover AI training under the meaning of "develop" business products. James Grimmelmann, a Cornell Law School professor, suggested the terms could allow Meta to build a comprehensive content archive. The News/Media Alliance sent a letter to Meta CEO Mark Zuckerberg, demanding revisions over concerns of "severe harm" to publishers.

The AKUSO take: This affects any business using Meta's platforms, especially media companies. Publishers are pushing back hard on what they see as a grab for content to train AI models; the delay may just be a pause before a larger fight.